So i was looking into this thought experiment and was wondering how far off we are (technologically) from being able to Matrix people or if it’s even really possible. My real question though, is “would an organism be able to tap into another’s brain and ‘feed’ it information/stimuli so that the affected would experience a simulated reality?” I suppose this hypothetical life form would also effectively be a mind reader so that it could have the false reality change/react to the affected’s actions in the sim.
I think it’s a safe guess to say “yes” it’s possible.
But if you pretend that this capability is analogous to building a 747, we have probably successfully created the windshield wipers and the seats in row 35, and that’s about it.
I don’t think it’s possible, even in theory. You are asking to simulate reality completely convincingly for an organ (the human brain) which is superbly adapted to comprehending reality very, very fast, almost on a real-time basis. I don’t think you can simulate reality in full detail on a real-time basis. You are essentially asking for a processor that works faster than events happen, which seems logically contradictory to me.
You can certainly do well, given the brain’s tendency to take shortcuts – to remember the gist of things, rather than full detail. Perhaps that is enough, in the end, to make it possible. A sufficient understanding of how the brain processes sensory data, and the short-cuts it takes, will allow you to simulate reality in much less detail than it actually has, and then you can do that in real time. But I’m skeptical. If the brain could possible be any faster and more efficient at comprehending sensory data, evolution would have made it so. It’s reasonable to conclude there is, and can be, no machine faster and more reliable at comprehending reality than the human brain.
You could also try to get around the problem by precomputing as much of the experience as possible, the way Hollywood spends months and years rendering computer-graphics that take seconds to minutes to show in a movie. But you can only do this for parts of the simulation that the actions of the brain cannot affect. Basically, you’re building “cut scenes” for your “video game.” As anyone who’s played video games knows, that’s hard to do fluidly, and in a way that doesn’t reveal the constraints of the experience being a game.
We already are brains in vats, with information about the outside world pumped in through our senses, which we can only assume represents an external reality and is not just some movie or simulation.
So the question is, how close are scientists to transmitting arbitrary information over our nerves and into our brains? The answer is not very. How close are scientists to even being able to keep a brain alive without a body to house it? Again, not very.
But I don’t see any reason to believe it isn’t possible, because that’s what our bodies do already. It’s just a matter of learning enough about how biology does it so we can copy, and maybe improve.
“Full detail” is not possible, nor necessary. Our senses have a certain bandwidth, that’s all the information you need. And you certainly don’t need to simulate all of reality. Only the parts they experience.
Also, people live every day without certain senses, so there’s no reason you couldn’t activate certain senses and not others. Meaning, you can send the vat brain audio information without corresponding visual information, save on bandwidth, and it would just perceive it was blind. Or not perceive it at all. Do blind children know they’re blind without being told?
I think the “convincing simulation” part will be a piece of cake compared to the “keep brain alive” and “transmit information to it” parts. We already have video games that are pretty convincing immersive experiences. Some people spend more time in World of Warcraft land than “real reality” land.
The problem isn’t intractable. You would need a immune system growing in a synthetic environment. That has been done experimentally. You need a pump for the heart and gas exchange membrane for the lungs., you need the liver cells growing in some type of synthetic matrix. That has already also been demonstrated. As for input, that is a lot easier than it sounds. You would use electrode grids wired to the output nerves, such as the motor homunculus. You use wires to the input nerves such as the sensory homunculus. You would do synthetic vision by wire to the visual cortex, and wiring to auditory nerves similar to existing implants. The problem, is that all of these things have been demonstrated very poorly in past experiments. Putting it altogether at today’s levels of development would be virtually certain
Also, starting the brain up without a body means you have to duplicate signals to he reticular activating system, and there are various other nasty problems that are currently unknown. You would not need a virtual environment, you could use translated live signals from cameras and microphones. TLDR : I think if solving this problem were an urgent national priority, and you had a giant budget and enough time, the problems could be solved.
ShaggyDredlocks We regularly have a GQ thread asking if we can keep a brain alive in a jar just yest. One user even had the sigline
I agree with the consensus that we’ll be able to do it eventually.
Re Detail And Bandwidth
If the simulated reality is only being fed to one brain, all that needs to be simulated at any one time is what that person can perceive. For example, I’m currently at home in my living room. If I were the brain in question, not the rest of the universe, not the rest of the galaxy, not the rest of the planet, not the rest of the country, heck not even the rest of my apartment building needs to be simulated. The television doesn’t need to be simulated unless I turn it on. Visually, you only need to simulate what I’m looking at- and then with much less detail than most people think except in the center of vision.
It also seems to me that you could make errors in the simulated reality that would go unnoticed. If you messed up some calculation involving the speed of light, how would the brain know? They have access only to the information you allow.
Firstly, thanks for all the feedback everyone.
But just to reiterate my intended question, because I stupidly sandwiched it in the middle of my post:“would an organism be able to tap into another’s brain and ‘feed’ it information/stimuli so that the affected would experience a simulated reality?”
To clarify, the affected individual is completely alive and well. So keeping the brain alive doesn’t factor into the scenario. I only brought up the brain-jar because its a well known sci-fi trope and comparable to my hypothetical life form.
Sorry I worded my OP sloppily.
We sort of already do. Encompassing virtual reality with visual auditory and even tactile feedback that gets brains to experience something that is not really there. If it can be done by stimulating the retina and the hair cells of the cochlea (with light and sound stimuli) then it can be done by stimulating the nerves if such was desired. Not an easy step but a very comprehensible one. Let the innate wiring figure out what neurons then get stimulated to what magnitude in what order from there with maybe a little ramp up of the motivation centers, like amygdala for fear response … to set the mood only.
Carl Pham’s aside and Dr Cube’s comments are cogent. The brain is wired to see signal and very forgiving of noise. If there is a pattern that can be imagined out of the inut a brain will tend to see it and fill in details that are not really there but which will be percieved. We can’t help it. That’s the basis of perceptual illusions and would work here as well.
Even more than this, how often are we wrong about something we think we see? You see a weird shadow in the living room and your brain says “Intruder!” and then you take a step forward and you say “Nope, just some ruffled curtains.” Or think about how much eyewitnesses can disagree about the events involved in incidents like car accidents. People perceive/remember things incorrectly all the time. As long as the simulation can correct flawed inputs quickly enough, the fact that the simulation might be wrong occasionally won’t be any worse than the brain interpreting that information.
Either that or the brains will continue to make up something to explain what they thought they saw. All those bigfoot stories are fabrications, but they may not be consciously fabricated.
This is a common, but highly questionable, philosophical assumption that likes to masquerade as scientific fact. Basically, it is Descartes’ theory of the mind-brain with the immaterial soul, that Descartes had sitting at the center of the brain, taken out, and replaced by the desperate pretense (accompanied by a lot of handwaving whenever the issue is pressed) that the rest of the theory did not entirely depend on its being there.
It is true that a lot of modern neuroscientists still conceive of the brain’s function in terms of this hoary old Cartesian philosophical framework, but that is becoming less and less true as it comes to be realized that there are alternatives that accord better with the actual scientifically established facts. You might want to check out the, by now, very large literature on embodied cognition, situated cognition, active perception, enactive cognition, intrinsic brain activity, etc., all of which challenge the Cartesian notion of the brain as a passive recipient of experience.
As for whether we are, technologically, anywhere near being able to create a brain in a vat, as in the traditional philosophical though experiment, the answer is that that we are a very long way off indeed, to the extent that, once you understand the issues in some detail, it becomes less than clear that it is even possible in principle. Here is an article [PDF] setting out some of the real issues. The brain’s connection to the rest of the body is of enormous bandwidth, and of multiple types. It is not just a matter of a lot of nerve cables (though there are a fuckton of them). And, of course, the brain’s connection to the body is bi-directional. The brain does not just take signals in, it sends them out too, and these outgoing signals, amongst other things, cause the body to move, which, in turn, changes the inputs. The evidence suggests, in fact, that a very large proportion of the brain’s activity is concerned with moving the body (or parts of it, such as the eyes - it has been plausibly claimed that most human behaviors are eye movements - or even less noticeable parts, such as the outer hair cells of the cochlea of the inner ear) in order to change and modulate the incoming signals, mostly in such a way as to make them more informative.
The situation depicted in The Matrix, incidentally, is importantly different from the idea of a brain in a vat, though the fact that so many people so readily conflate them is evidence, I think, of just how little thought has usually been given to the real complexities of the envisaged situations. Matrix brains are still in their original bodies, which makes all the complex problems surrounding the issue of the brain’s very rich, multifarious interconnectedness with the body, which loom very large for the vat scenario, quite irrelevant. It does not, however, solve the problem of the bi-directionality of the brain-body information flow, and the fact that brain is constantly, on a fraction-of-second timescale moving and otherwise modulating the sense organs in order to change the incoming signals and obtain the information it wants. Even in the Matrix, the machine cannot just be concerned with feeding the brain with a certain input, like a 3-D movie. It has to be concerned with how inputs would change when the brain outputs signals that would in normal circumstances, cause the body (maybe just the eyes, or fingers, or other unobtrusive part, to move (or do other things that modulate the input signals - it is not just movement, sensitivity of sense organs can be modulated and tuned on a fine-scale level, for instance). When you take all of this into account, it becomes plausible that any computer capable of dealing with all this, and thus producing a fully deceptive “virtual reality”, would have to be as large and complex as the external world itself (which is, of course, impossible, because, ex hypothesis, the would includes the computer that is running the Matrix simulation).
[Actually existent virtual reality systems, apart from falling well short of being able to simulate a truly deceptive world, differ in other important ways from the Matrix, let alone the brain-in-vat scenarios. They do not attempt to feed information straight into your brain, but use your regular sense organs that remain subject to the normal movements and oterh modulations that the brain imposes on them. The eyes, for instance, still move freely, and are presented with genuine visual information on a screen, from which they can freely select, in the normal way, by moving to look at different parts of it.]